- frak — transforms collections of strings into regular expressions for matching those strings. The primary goal of this library is to generate regular expressions from a known set of inputs which avoid backtracking as much as possible.
- The Boolean Trap — crappy APIs where true/false don’t mean what they seem. None of this is new, but every generation has to learn it anew. (via Pete Warden)
- Gumbo — open source pure C HTML5 parser.
- All Those People With Cheap Android Phones Have Started Buying Apps (Quartz) — revenue generated by the Google Play Store, from which many Android users get their apps, has grown 67% over the past half-year. By contrast, Apple’s App Store revenue grew 15%, according to Distimo estimates, which cover the top 18 countries. That sounds less impressive if you consider that the App Store brought in twice as much revenue in absolute terms. But the numbers are shifting fast. In February, the App Store was generating three times as much revenue, and last November it out-earned Google Play by a factor of four. Google is gaining.
"app store" entries
Getting apps into the store is a non-deterministic process
One of the major topics of my Enterprise iOS book is how to plan release schedules around Apple’s peril-filled submission process. I don’t think you can count yourself a truly bloodied iOS dev until you’ve gotten your first rejection notice from iTunes Connect, especially under deadline pressure.
Traditionally, the major reasons that applications would bounce is that the developer had been a Bad Person. They had grossly abused the Human Interface standards, or had a flakey app that crashed when the tester fired it up, or used undocumented internal system calls. In most cases, the rejection could have been anticipated if the developer had done his homework. There were occasional apps that got rejected for bizarre reasons, such as perceived adult content, or because of some secret Apple agenda, but they were the rare exception. If you followed the rules, your app would get in the store.
Robo Regex, Crappy Code, HTML5 Parsing, and Play v App Numbers
The App Store model has increased the uncertainty of the software release process
The recent unavailability of the Apple Developer’s Portal just underscores how increasingly dependent developers have become on third parties during the software lifecycle. For those who are not following the fun and games, the developer.apple.com sites, which include much of the functionality needed to develop Mac and iOS applications, has been unavailable for more than a week as of this writing. Although iTunes Connect, the portal used to actually deploy apps to the App Stores, has remained available, the remainder of the site territory has been off-limits. This is all thanks to a security intrusion (evidently by an over-zealous researcher.)
The App Store model has fundamentally changed how software is distributed, mostly for the better (IMHO), but it has also removed some of the control of the release process from the hands of the developers and companies they work for. As I have spelled out previously in my book on iOS enterprise development, the fact that Apple has the final say on if and when software goes into the store has required more conservative release timelines. If you want to release on the first of September, you need to count back at least two weeks for “gold master”, because you need to upload the app, potentially go through a round of rejection from Apple, and then upload a fixed version.
Android apps don’t suffer from this lag, because most of the Android stores don’t do any significant checking of the applications uploaded to them. The Devil’s Deal that Apple developers have made with Apple is that in return for the longer wait time to get apps in the store (and having to follow Apple’s rules), they get a de facto seal of approval from Apple. In other words, it is assumed that apps in the iTunes store are more stringently policed and less likely to crash or do harm (deliberately or else-wise.)
The current downtime has brought that deal into question, however. Suddenly, developers who need new provisioning certificates, passbook certificates, or push notification certificates find themselves with nowhere to go. Even if iTunes Connect is available, it doesn’t do you any good if you can’t get a distribution certificate to sign your app for the store. I’m sure that there are developers at this moment who have had their finely tuned release strategies thrown into disarray by the in-availability of the developer portal.
Being essentially at the mercy of Apple’s whims (or Google’s, for that matter) can’t be a pleasant sensation for a company or individual trying to get a new piece of software out the door. The question that the developer community will have to answer is if the benefits of the App Store model make it worth the hassles, in the long run.
Don't Pay Developers, Teaching Programming, Second Android Screens, and Democracy
- Paying for Developers is a Bad Idea (Charlie Kindel) — The companies that make the most profit are those who build virtuous platform cycles. There are no proof points in history of virtuous platform cycles being created when the platform provider incents developers to target the platform by paying them. Paying developers to target your platform is a sign of desperation. Doing so means developers have no skin in the game. A platform where developers do not have skin in the game is artificially propped up and will not succeed in the long run. A thesis illustrated with his experience at Microsoft.
- Learnable Programming (Bret Victor) — deconstructs Khan Academy’s coding learning environment, and explains Victor’s take on learning to program. A good system is designed to encourage particular ways of thinking, with all features carefully and cohesively designed around that purpose. This essay will present many features! The trick is to see through them — to see the underlying design principles that they represent, and understand how these principles enable the programmer to think. (via Layton Duncan)
- Tablet as External Display for Android Smartphones — new app, in beta, letting you remote-control via a tablet. (via Tab Times)
- Clay Shirky: How The Internet Will (One Day) Transform Government (TED Talk) — There’s no democracy worth the name that doesn’t have a transparency move, but transparency is openness in only one direction, and being given a dashboard without a steering wheel has never been the core promise a democracy makes to its citizens.
Three new battles to watch as the mobile hardware gap closes.
The iPhone 5 may or may not be the most beautiful handheld device, but it barely matters anymore. Competitors have rendered its beauty and craftsmanship irrelevant. Amazon has received the message and responded with its latest set of tablets, and Google has responded with the Motorola Droids and the Nexus 7. These devices now have sufficient quality in their materials, specs, and base operating systems so that they can make any consumer happy. So if hardware is a toss up, where will the next battles be fought?
The answer: developers, integration, and discovery.
First, the very best developers will build apps that tap key trends: improved camera quality is making real-world text and face recognition more possible, geofencing data stores are making proximity–based apps more possible, and despite Steve Jobs’ assertion that God gave us 10 styli, there’s clearly a host of applications that are benefiting from pressure-sensitivity and pens. The level to which Apple and Google embrace these new technologies and extend the current state of the art in voice and gesture recognition will factor heavily into the quality and emergence of new applications. In addition, the extent to which Apple and Google can expose these new technologies — like NFC or always-on Glass cameras in Google’s case — will provide an advantage to developers.
App Economics, Kickstarter Lessons, CSV Code, and Science Magic
- 5 Is The New 10 — I have limited sympathy for the “app developers can’t predict their fortunes” complaint: creative arts have always been long tail hit-based businesses, possibly because hits have a large random component.
- Lessons for Kickstarter Creators (Mat Howie) — great case study of a disastrous KS project. Preparation, research, and comms are what let this one down. (via Mat Howie
- CSV Kit — commandline tools for working with CSV files. (via Hadley Wickham)
- Science of Magic — magic tricks which help you teach students how to apply the scientific method. Magic and science both built off flaws in human perception and intuition: science tries to avoid them, magic to exploit them. (via Maria Popova)
A good hacker is hard to find, a cheap data center is hard to get to, and the app store model is hard to ignore
An argument for the value of highly productive programmers, datacenters head for the country to save a few bucks, and the app store model seems to be taking over the industry, and not just for mobile.
DIY Bio Hardware, App Store Numbers, Open Hardware Repository, and Science Startups
- OpenPCR Shipping — A PCR machine is basically a copy machine for DNA. It is essential for most work with DNA, things like exposing fraud at a sushi restaurant, diagnosing diseases including HIV and H1N1, or exploring your own genome. The guy who discovered the PCR process earned a Nobel Prize in 1993, and OpenPCR is now the first open source PCR machine. The price of a traditional PCR machine is around $3,000. This one is $512 and would go well with Ben Krasnow’s Scanning Electron Microscope. Biological tools get closer to hobbyist/hacker prices. (via Gabriella Coleman)
- Apple App Store Figures (Fast Company) — 1 billion apps in a month, 200M iOS users, $2.5B revshare to developers so far (implying a further $5.8B revenue kept by Apple). Another reminder of the astonishing money to be made by riding the mainstreaming of tech: as we move from dumb phones to smart phones, the market for Apple’s products and App Store sales will continue to rise. We’re not at the fighting-for-market-share stage yet, it’s still in the boom. (via Stephen Walli)
- Open Hardware Repository — open source digital hardware projects, such as a tool for generating VHDL/Verilog cores which implement Wishbone bus slaves with certain registers, memory blocks, FIFOs and interrupts. CERN just approved an open license for hardware designs. (via CERN)
- Wingu — SaaS startup to help scientists manage, analyze, and share data. Recently invested by Google, it’s one of several startups for scientists, such as Macmillan’s Digital Science which is run by Timo Hannay who is one of the convenors of Science Foo Camp. (via Alex Butler)